Per-word pricing assumes every word takes the same effort. AI drafts broke that assumption. Recorded at LocWorld 56 Vancouver, this episode of The Agile Localization Podcast brings host Stefan Huyghe together with Masahito Sugahara, Technical Program Manager for Internationalization at Atlassian, to share a pricing model that ties translation discounts to how much linguists actually edit AI output.
The views, thoughts, and opinions expressed in this podcast are solely that of the speaker and do not reflect the views, opinions, policies, or position of Crowdin.
Per-word pricing assumes every word takes the same effort. AI drafts broke that assumption. Recorded at LocWorld 56 Vancouver, this episode of The Agile Localization Podcast brings host Stefan Huyghe together with Masahito Sugahara, Technical Program Manager for Internationalization at Atlassian, to share a pricing model that ties translation discounts to how much linguists actually edit AI output.
What You’ll Learn
- Why surging AI-driven product content made traditional translation budgets unsustainable
- How fixing broken plurals, placeholders, and hard-coded strings at the source changes what translators see
- Why a quarterly cadence smooths out the gap between edit counts and real effort
- How quarterly, per-language discount tiers based on TER work in practice
- What smaller teams and LSPs need before proposing a similar model
- Why rate floors are essential to protect linguists under an edit-based model
Masahito Sugahara is the Technical Program Manager for Internationalization at Atlassian. He works across product, engineering, and vendor teams to keep Atlassian's global content flowing as volumes grow, from fixing internationalization issues in source code to rolling out AI translation at scale. After two years of piloting and quality testing, Masahito helped design a dynamic pricing framework that links vendor discounts to Translation Edit Rate scores, with MQM-based quality monitoring and rate floors built in.
If you enjoyed this episode, make sure to subscribe, rate and review on Apple Podcasts, Spotify and YouTube Podcasts, instructions on how to do this are
here.
The views, thoughts, and opinions expressed in this podcast are solely that of the speaker and do not reflect the views, opinions, policies, or position of Crowdin.
Episode Resources: